LAB #4881 NEUTRAL: LEDGER BENCH: Apply filter: NZDUSD LONG @ NY_OPEN with 6 conf — improves win rate from 31.9% to 82.8 — E8 Intelligence Research

IDEA: Adjustment made to the live model — verify against the backtests: Apply filter: NZDUSD LONG @ NY_OPEN with 6 conf — improves win rate from 31.9% to 82.8% (+50.9pp, n=29) SAME-WINDOW EFFECT: The filter would not have touched a single trade in the live window above — every logged position is either SHORT NZDUSD (the filter is LONG-only) or a different pair entirely, so the affected-trade count is zero and the win-rate/PnL delta over this window is exactly nil. Worse, the one NZDUSD trade in the window (SHORT @ 0.58415, 2026-09-09) is the opposite direction of the proposed filter, meaning the rule as stated would have blocked nothing and added nothing. The +50.9pp claim is therefore untestable against this window and cannot be corroborated by the live ledger. OVER-TIME PROJECTION: A 82.8% win rate on n=29 is a sample so small that the 95% confidence interval spans roughly 65–93%, and the baseline it's compared against (31.9%) is drawn from a different, much larger pool — the compa Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com